83 VII.-On a- nnd p-Pl~e?z.antl~rene-ct-crl)o;e!jlic Acids with Remarks o u tlie Constitution of Plw.nantlwene. By FRAKCIS R. JAPP M.A. Ph.D. Assista,nt in the Chemical Research Laboratory Science Schools South Kensingtou. I HAVE already described ( B e y . 10 1661) in conjunction with Dr. Schultz a phenaathrene-carboxylic acid melting a t 260". I have since prepared this acid in greater quantity and having thus been enabled to purify it more thoroughly find t h a t it melts a t 266". I now pro-pose to designate it as a-phenanthrene-carboxylic acid in order to dis-tinguish i t from the isomeride described for the first time in the present paper. a-Phenant hrene- carboxylic acid was ob tained from crjstallisecl calcium phenanthrene-sulphonate by converting the lather into the sodium salt distilling the dry sodium salt with potassic ferrxpanide, and saponifying the nitrile thus obtained.I n preparing the calciiim phenanthrcae-sulphonate there was always ;t relatively small yield of crystallisable salt and a considerable quan-tity of dark-coloured syrupy mother-liquor. This latter had been f'roin time to time reserved for investigation. As it seemed possible that this mother-liquor might contain an isomeric calcium phenanthrene-sulphonate I determined to subject it to the same processes which had yielded a pheiiantlirene-carbox~-lic acid from the crystallised salt hoping that the corresponding pliena n-threne-carboxylic acid might be more easily purified than the syrupy On adding sodium carbonate until the solution was all ;a I' me a copious precipitate of calcium carbonate was formed denoting the presence of a very soluble calcium salt.The filtered solution of the sodium salt was evapora>ted to dryness and the dried salt was mixed with one and a half times its weight of potassium ferrocpnide and distilled in portions of 100 grams a t a time from a flat copper retort. The nitxile thus obtained was a yellow transparent viscid liquid, which after standing for some days assumed a buttery consistency, probably owing to the presence of regenerated phenanthretie. It mas saponified with alcoholic potash and required boiling for over lot3 hours with inverted condenser before ammonia ceased to be evolved. The alcoliol was then distilled off and the contents of the flask after digesting with water were diluted and filtered.OH acidifj-illg with hydrochloric acid the new acid separated out as a dirty white floccn-lent precipitate. The yield of crude acid from 2 kilos. of coilrmercial phenanthrene x-as only 80 grams, s ulpIlor1 at e . G 8 4 JAPP ON a- AND P-P€IEN_~NTIIRESE-C~~RBOSPLIC ACIDS. The purification of this acid presented considerable difficulties. It dissolved in ammonia yielding a brown liquid. On adcling an excess of barium chloride the barium salt was precipitated carrying witli it, most of the resinous colouring matters. On boiling the barium salt went partly into soluticn yielding a nearly colonrless liquid which was filtered hot from the impurities. The residue had to he repentedly extracted with boiling water as the impnritics enclosed the salt ant1 rendered its solution difficult.Subsequent experience with the barimn salt has shown that a re-peated recrystallisation of this salt would probably have offered it good means of purifying the aci(1 bnt a t this stage the salt did not seem to promise well and its use was abandoned. After a trial of several salts the sodium salt which ~rystallised in long rhomboidal lamin= with a slight satiny lustre mas selected. The snlt was recrys-tallisecl serernl times until it was quite colo~-rrless. On adcling hydro-chloric acid to tile solution /3-yhennnthrene-carboxylic acid was ob-tained as a white flocculent precipitate. 6-Phenanthrene-cnrboxylic acitl is soluble in alcohol ether nncl glacial acetic acid almost insoluble in water. From a hot saturateti acetic acid solution it crj-stallises in stellate groiip of colonrless needles.It sublimes in fern-slinpetl leaves exactly resembling those of t>he a-acid. Analysis yielded the following results :-I. 0.2268 gram gRve 0.6765 gram carbonic anhFclriile and 0.0938 gram water. 11. 0.2837 gram gave 0.8453 gram carbonic anhydride and 0.1151 grain water. It melts a t 250-2552'. Pound. -\-Calculated for CI,KloO2. I. 11. C, . . . . . . . . . . 180 81.08 81.36 81.23 HI . . . . . . . . . . 10 4.50 4.59 4-51 0,. . . . . . . . . . . . s2 14-42 (14.05) (14.26) 222 lu0fw 100.00 100.00 A determination of water of crystallisation and sodium in the sodium s n l t above referred t o g3ve the following results :-0.3782 gram (of a preparation which had been exposed to the air for a long time) lost on heating to 140" 0.0986 gram and the remain-ing 0.2796 gram anhydrous salt when treated with concentrated sul-phuric acid and heated to redness gave 0.0798 gram sodium sulphate.The formula Cl1H,.CO2Na + 5H20 requires 26.96 per cent. H,O. Fouiid 26.07. The forniiila ClpE,.COzNa (anh-ylrous) requires 9-43 per cent. Na. Found 9.25 J A P P OX tl- ,4SD P-PHESASTIIRESE-CXnEOSTLIC ACIDS. 85 The barium salt was obtained as a crystalline precipitate when lmriurn chloride was added t o a solution of the sodium salt. It mas 1)urified by recrystallising from a hot aqueous solution and formed colourleas branched needles w hicli under the microscope appeared as long rectangular laminze. These lamiwe were verj brittle and pos-bessed a cleavage parallel to the right angled terniiristion.Water of crystallisation and barium were deteriniiied with the following re-sults :-I. 0.3763 gram air-dried salt lost on heating to 140" 0-0582 gram, and the remaining 0.3185 gram anh~drous salt gave 0.1276 gram barium sulphate. 11. 0.3185 gram lost on heating to 140" 0.0490 gram and the re-niaiiiiug 0.2695 gram anhydrous salt gave 0.1070 gram barium sulphate. The formula (CI4H,.CO2),Ba + GH,O requires 15.71 per cent. HzO. Found-1. 15.45; 11. 15-38. The formula ( C,,HT,.CO,),Ba (anhydrous) requires 23-66 per cent. Ba. Pound-I. 23.57; 11. 23-& D isti1 ILI t ion of P-l'henail tli retie- c-nrbozy lic Acid with Sodu-1 i me. I K ~ order t o prove that the acid was really a derivative of phenaii-tlirenc 8 grams of the sodium salt were distilled with soda-lime.The phenaiithene thus obtaiiied was identitied by its melting point by that of its picric acid double compound a i d lastly by tliat of the yuiriorie prepared f rcmi it by oxidation. The other physical propertics of these compounds agreed perfectly with those of the phenanthreiie compounds in question. Furthermore t8he quinone dissolved without residue iii a solution of acid sodium sulpliite. OxidutLutL o j p- Pl~e?.ln,ithl.etLe-cu~.~~~~li~ Acid. 1 grain of the acid was oxidised with about twice its weight of chromic anhydride in acetic acid solution. On distilling oB the excess o f acetic acid and diluting with water a substance was precipitated in orange-yvellow needlcs froin which a solution of sodium carbonate extracted R small quantity of unoxidised phennnthrene-carboxylic acid.'l'he residue dissolved almost entirely in a solutisn of acid sodium sul-phite. The pure substance reprecipitutzcl from this solution was cry+ tallised once from glacial acetic acid aid was thus obtained in orange-~ d l o w needles melting a t 204-2U4*5'. (31. p. of yllenanthrene-quinone given by Fittig a t 198" found b j myself 011 a very pure specimen as high as 206O.) Analysis yielded the following results :-0.1924 gram gave 0.566% gram carbonic a n h y d d e and 0.0711 gram l v n t x 86 JAPP ON C( - AND P-PHENANTHRENE-CARBOXPLIC ACIDS. Calculated for CI4HSO2. Found. Cli 168 80.77 80.51 Hs . . . . . . . . . . 8 3.85 4.10 0 . . . . . . . . . .. . 32 15-38 (15-39) 208 100.00 lOC!.OO The substance was therefore pure pheilanthrene-quinone. Tabulated Co?npariso?t oj' a- and P-Phermr th-ene-czcrboxylic Acids. The following table contains the result of a comparison of the two phenanthrene-carboxylic acids and their sodium and barium salts. Most of the crystallographical characteristics here described can he per-ceived only with the aid of a microscope :-A cici. . Sodium salt. . Barium salt u- Acid. ~~~ Crystallises from hot glilcid acetic acid in colourless curved blades with pamllel edges and a rectangular termination. This currature is very characteristic. Sublimes in fern-shaped leares. M. p. 266". (The melting point 260O was given in the first p p e r . ) of colourless pointed blades.100 parts of water at 20" dis-solves 6.8 parts of the anhydrous salt. Almost indefinitely sdu-ble in boiling water. C(IIHS.C(OQNa + 4H20. Tufts (C14110CO~),Ba + '/H,O. Colour-less long needless of eutraoidi-nary fineness mid flexibility, radiating from one point to form large balls or tufts. Under the microscope .a t a n g l d mass of these flexible needles has the appearance of vegetable fibre. 100 parts of water clissolve-p-Acid. Crptallises from hot glacid acetic acid in sttellate groups of colourless st8i*aiglit pointcd needles. S nblimes in fern-shaped leares. 31. p. 250-252O. C,lIf~.CO,Pu'a + 5&0. Colour-less rhoinboidal laminq with a slightlg satiny lustre. 100 parts of water a t 20" dissolve 6.2 parts of the anhydrous salt.Almost indefinitely soluble in boiling water. (C,JI,.CO,)nBa + GH,O. Co-loarless long brittle rect:m-gulsr Isminre united in a rami-form erystallisntion. 100 parts of water dissohe-At 20'. . -27 parts of anhy-, 100" 3.70) drous salt. !K% eoretical Considerations. It has been shown (Jappand Schultz loc. tit.) that a-phenanthrene-carboxylic acid yields on oxidation phenant hrene-quinone-carboxg lie acid the carboxyl group remaining intact and that a-phennnthrene-ca,rboxylic acid has therefore the formula JAPP 0s Z- ASD P-PI-IE-U-4~'THRESE-C:1RI305YLIC ACIDS. 87 C0.OH I C,H,-CH, 1 II C,H,-CH the exact position of the carboxyl group in the benzene nucleus re-maining undetermined. A phenanthrene-carboxylic acid 117 hich on oxidation yields plienan-threne-quinone with elimination of the carboxyl group must contRin tlhis group attached to one of those carbon atoms which in the quinone are united with quinonic oxygen.This would give-C,H C- C 0. OH CGHA-CH I I1 as the constitutional formda of P-phemnthrene-carboxylic acid,* and-C,HA-C-SO,.OH as the formula of the sulphonic acid from which i t is derived sup-posing no migration of atoms within the molecule to have taken place during the distillation with potassic ferrocya,nide.+ The formation of this sulphonic acid by the direct action oficon-centrated sulphuric acid on phenanthrene seems to me to cast fresh light on the constitution of phenanthrene and indirect,ly on that of the benzene nucleus. The graphic formula of phenanthrene expressed in terms of KekulA's benzene theory is-* Licbernianii and rom Rath (Ber.8,248) seem to hare observed the pesence of tliis acid as an impurity in their anthracene-carboqlic acid. They mention that 8 specimen of the latter acid prepared from impure nnthrxcene yielded on oxidation, in addition to anthraquinone-carboxylic acid phenanthene-quinone. -t As this point was of importance the following experiment ~ v a s undertaken to decide it :-A portion of the very soluble sodiuni phenanthrene-swlplionate from which the P-phenanthrene-carboxjlic acid H'BS prepared was dissolved in water anJ oxiclised with a mixture of potassium clichromatc and sulphuric acid. Carboni 8 8 JAPP ON U- AXD P-PHE~X~THRESE-C~RCOSTL~C ACIDS. a plain hexagon being employed t o sFmbolise the benzene nucleus iri ordcr to avoid introducing the conip1ic;Ltcd question of alternate double and single bonds t o which LoweT-cr reference will be made further on.Grnebe’s synthesis of this compound from stilbene-symnietrical cliphenyl-ethylene-shows it to be diplienylene-ethylene. Schultz ( B e r . , 111 215) on the one Iiand and Anscliuetz and Japp (ihid. 11 211), ( a i l the other showed by two iridepeiiclent methods that phenanthrene is a diortho-compound arid that i t must therefore be regarded as symmetrical dioytho-dipheuyleire-etliFlene as expressed in the above graphic formula. Many of the chemical reactions of phcnantLrerie seem to indicate that the dyad ethylene residue -CH=CH- i n tlic diphenylene-utliylcne forms pait of an aromatic nucleus.On oxidation the two atonis of hydrogen are replaced by two atoms of oxygen (not by one), a quiuone being formed. This quinone 011 reduction yields a hydro-quinone containing the group -C (OH)=C( OH)- and dissolving in caustic alkalis by virtue of these plicnylic hydroxyl groups. Mono-bromphenanthrene in which the s u h t itution of bromine takes place in the -CH=CH- group-tlie coniponnd yielding on oxidation phenanthrcne-quinone tvitli elirriiiiatioii of the bromine atom-may as Aiischuetz has shown (BcI-. 31 1217) be heated with strong alcoholic potash to 170” without undwgoirig change. Aiiscliuetx points out the bearings of this fact and the importance of a proof of the aromatic chaiacter of this portion of the phenanthrene molecule. Indeed, hen one considers tlie relatire mobility of the bromine atom i n monobromctbylene and mono~~urnstilberle lien these compounds are lreatcd with alcoholic potash this stability of rnouobromplienaiithreiie innst be allowed to furnish very strong cvitlence that the group - - C H ~ C H - forms part of an aroinatic iiucleus.A fresh picce of evidence of a similar cliaimAer may be found in the :ibove-described direct formation froni pLenniitbrene atid sulpburic mid of a sulplio!iic acid in whieli tlie sulyholiic group replaces one of the hydrogen atoms of the - - C H ~ C H - group. A direct snlphoiiation of a liyclroc~arbon has as yet h e n observed only in the ttroniatic nucleus. I n the fatt)y series the presencc of an elcctro-nega-tivc grmp (CN CO.OII &c.) is necessary in order that direct sul-1)hoiiation may take place.r i 1 licse reactioiis are without exception spccificaliy aromatic in anhydride 11 as e%oivctl niid n n or:iiigt.-j t.110~ precipitate I\ as formcd ~vliich after 1:urifjing in the uiaiincr already desct ibcd (see “ Osidatioil of P-I’henantlireiie-earbox>lic Acid,” 1). S;) crj stalliscd in ~ieedloa fusing betwrcri 198‘ and 200° and exhibiting all the other proprties of ~)licii:~ritlireiie-q~~ixi~xi~~. The presence of’ a I’~ienantlireiic-sulplioiiic acid of the a h \ e forriiulil as t1iu.i proved character and taken together form a strong cumulative proof t h a t the -CHECH- group belongs to a hcnzerie nucleus. In the synthesis of pliensiithrene from’ syminctrical diphenyl-ethj-lene we have therefore constructed a benzene nucleus from the ctliplene residue -CHXCH- on the one hand (which as long as i t l>xisted in diphenyl-ethylene showed by its e:itire behavionr that it (lid not belong to a benzene nucleus) and on tlie other two pairs of (.arkion atoms each of whicli pairs WRS contained i n the ortho- position I)y a separate already existing nucleus :-C / c II I - CH II CH I c I’iienantllrene therefore collsists 02 three Leiizene nuclei one of which shares f o u r adjacent carbon atoms wit11 the two others-one ortho-pair with each.Phenanthrene niay thus be derived from lmphthalene by a repetition of the process by wliich thc latter hydro-carbon is derived from benzene-a suggestion thrown out by Graebe ( A m . C l ~ e n ~ . P J i a ~ i i ~ . 167 133) in his memoir on plieuant>hrene.The portion of tlie liydrocarbon to the right o€ thc dotted line i n tlie above graphic formula is the dyad group to which reference has been iuatle. Ldenburg has shown (T‘lLcui-ie der amuatischeii V w b i t d u ? L g e i ~ ) that c d j 1wo graphic formula satisfy the coriditioris required by a benzcne i~ucleus-his own “ prismatic ” formula aid Kekulh’s hexagon with alternate double and single boiids or wliat is practically identical witli this latter Lothar hleyer’s rriodified Kekul&’s liexagon wit11 six free afinitics. I f the central nucleus in phenanthrene is a benzene ~~ucleus then Ladenburg’s prism must be abandoned as i t is impos-bible to arrange three such prisms with four carbon atoms in tlie ortho-pof,ition in common so as to uatisfy the conditions required by t be phensiitlii-ene formula.* 1 It is equal!! iiupossihle t o express tlie naphthalene formula byr iiieans of Laden 90 EEDSON ON SOME DERIVATlVES OF As regards the formerly much vexed question of altcriiate double and single bonds in the benzene nucleus and whether the ortlio-com-pounds 1 2 and 1 6 ough% on such ft supposition to he identical, this difficulty could perhaps be best avoided by the adoption of Lothar Meyer’s hexagon with six free affinities.Carried to its logical conclusion this w o a l ~ l lead t o the universal substitution of free nEini-ties for double bonds as has been done before now. As long as wc have no physical conception of what a bond is this would involve no contradiction. All that a double bond represents is the potentiality of taking up two monad atoms or groups arid this is equally well accounted for by the assumption of free affinities. It is strange that in an opcn chain a double bond should represent the weakest part of the chain. A final proof of the aromatic nature of the central nucleus of pheiianthrene would be the preparation of naphtbalene o r one of its derivatives from this hydrocarbon. Experiments which I have made in this direction have as yet failed owing to the oxidisable character of the central group which causes it to be attacked sooner than the lateral nuclei even when the hydrogen atoms of one of the latter have been replaced by liydsoxyl groups in order to render the nncleus less stable. I t is evident from the graphic formula that five mono-substitution compounds of plienanthrene are theoretically possible